DNA repair plays a pivotal role in protection from carcinogenesis. The DNA repair protein 0(6)alkylguanine-DNA alkyltransferase protects cells from the mutagenic and cytotoxic DNA damage of agents such as nitrosoureas and nitrosamines which form DNA adducts at the O(6) position of guanine. From in vitro data in bacteria and cells, it is clear that lack of this protein increases the susceptibility to nitrosoureas and nitrosamines whereas cells with high levels of this DNA repair protein withstand much higher levels of exposure. Furthermore, in animals, a correlation exists between low levels of alkyltransferase and susceptibility to the toxic and carcinogenic effects of nitrosoureas. While such a relationship is hard to establish in humans, we have previously shown that bone marrow cells have low alkyltransferase which might explain the high incidence of nitrosourea induced leukemia among patients receiving these compounds. This proposal will test the hypothesis that increasing the expression of alkyltransferase will prevent an animal model of this leukemia. In this model, mice develop nitrosourea induced T cell lymphomas. We have successfully generated transgenic mice which express the human alkyltransferase gene MGMT in a number of tissues including thymus, spleen and bone marrow. The aims of this proposal are to 1) fully characterize the pattern of expression of the MGMT gene in transgenic mice by studying tissue and cellular expression of the mRNA in alkyltransferase protein. 2) Establish the capacity of the human MGMT gene to protect transgenic mice from the development of MNU induced lymphoma using a variety of dose schedules and genetic backgrounds. 3) Determine whether transgenic expression of human alkyltransferase results in rapid repair of O(6)methylguanine-DNA adducts and whether this rapid repair is uniform or heterogeneous in different tissues. 4) To determine whether transgenic expression of human alkyltransferase prevents mutational activation of k-ras oncogene in the thymus, the oncogene most frequently associated with MNU induced lymphomas and reduces the mutational frequency at the HPRT locus of T lymphocytes. This animal model will determine whether increased expression of a single DNA repair protein can prevent carcinogen induced cancer in mice. These studies have far reaching implications in the understanding of human carcinogenesis and in the use of genetic manipulations to prevent cancer in humans.